Investigating the urban air quality effects of cool walls and cool roofs in

Subscriber access provided by ALBRIGHT COLLEGE

Environmental Modeling

Investigating the urban air quality effects of cool walls and cool roofs in Southern California Jiachen Zhang, Yun Li, Wei Tao, Junfeng Liu, Ronnen M. Levinson, Arash Mohegh, and George Ban-Weiss Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.9b00626 • Publication Date (Web): 24 May 2019 Downloaded from http://pubs.acs.org on May 25, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36

Environmental Science & Technology

1

Investigating the urban air quality effects of cool walls and cool roofs in

2

Southern California

3

Jiachen Zhang1, Yun Li1, Wei Tao2, Junfeng Liu3, Ronnen Levinson4, Arash Mohegh1, George Ban-

4

Weiss1,*

5

1

6

CA 90089, USA

7

2 Multiphase

8

Mainz, Germany

9

3 Laboratory

Department of Civil and Environmental Engineering, University of Southern California, Los Angeles,

Chemistry Department, Max-Planck-Institute for Chemistry, Hahn-Meitner-Weg 1, 55128,

for Earth Surface Processes, College of Urban and Environmental Sciences, Peking

10

University, Beijing, 100871, China

11

4 Heat

12

*Correspondence to George Ban-Weiss, [email protected], 213-740-9124

Island Group, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA

13

14

Abstract

15

Solar reflective cool roofs and walls can be used to mitigate the urban heat island effect. While

16

many past studies have investigated the climate impacts of adopting cool surfaces, few studies

17

have investigated their effects on air pollution, especially on particulate matter (PM). This

18

research for the first time investigates the influence of widespread deployment of cool walls on

19

urban air pollutant concentrations, and systematically compares cool wall to cool roof effects.

20

Simulations using a coupled meteorology-chemistry model (WRF-Chem) for a representative

21

summertime period show that cool walls and roofs can reduce urban air temperatures, wind 1 ACS Paragon Plus Environment


22

speeds, and planetary boundary heights in the Los Angeles Basin. Consequently, increasing wall

23

(roof) albedo by 0.80, an upper bound scenario, leads to maximum daily 8-hour average ozone

24

concentration reductions of 0.35 (0.83) ppbv in Los Angeles County. However, cool walls

25

(roofs) increase daily average PM2.5 concentrations by 0.62 (0.85) μg m-3. We investigate the

26

competing processes driving changes in concentrations of speciated PM2.5. Increases in primary

27

PM (elemental carbon and primary organic aerosols) concentrations can be attributed to

28

reductions in ventilation of the Los Angeles Basin. Increases in concentrations of semi-volatile

29

species (e.g., nitrate) are mainly driven by increases in gas-to-particle conversion due to reduced

30

atmospheric temperatures.

31

32

1 Introduction

33

Urbanization is occurring at a fast pace around the world; global urban land area in 2030 is

34

projected to be up to triple that in 20001. Compared to rural areas with natural land cover, urban

35

areas contain more impervious surfaces that are made of solar absorptive and thermally massive

36

materials, such as asphalt concrete. Urban areas also contain less vegetation and thus reduced

37

evaporative cooling and shade cover. These differences in urban and natural land cover contribute

38

to the urban heat island (UHI) effect (i.e., cities being hotter than their surrounding rural areas)2,

39

which can, in turn, affect air pollutant concentrations. The air quality effects of urban land

40

expansion have been studied in previous research3–7, although only a few studies clearly explained

41

the mechanisms driving these effects8–11. Tao et al.8 suggested that with pollutant emissions held

42

constant, urbanization in eastern China would increase ozone concentrations from the surface to 4

2 ACS Paragon Plus Environment

Page 2 of 36

Page 3 of 36


43

km. However, it would also enhance turbulent mixing and vertical advection, therefore reducing

44

the concentrations of primary pollutants below 500 m.

45

While many studies have explored the air quality impacts of the UHI effect, fewer studies have

46

investigated how strategies that mitigate the UHI effect would influence urban air quality12–15. For

47

example, adopting solar reflective cool surfaces (roofs, walls, and pavements) increases city albedo

48

and the solar radiation reflected by cities, therefore reducing urban surface temperatures and near-

49

surface air temperatures16–22. However, adopting cool surfaces might change air quality in

50

unexpected ways. For primary pollutants (i.e., pollutants directly emitted to the atmosphere) such

51

as elemental carbon (EC), nitric oxide (NO), and carbon monoxide (CO), lower surface

52

temperatures in cities may suppress convection and therefore reduce atmospheric mixing heights

53

and vertical dispersion of pollutants, leading to increases in pollutant concentrations near the

54

ground23. Changes in horizontal temperature distributions can also influence wind speed and

55

direction, affecting the horizontal transport and distribution of pollutants. For secondary pollutants

56

(i.e., pollutants formed in the atmosphere from primary pollutants), in addition to the previously

57

mentioned changes in transport and dispersion of pollutants and their precursors, pollutant

58

concentrations can also be influenced by temperature dependent chemical reactions, phase-

59

partitioning, and emissions. Tropospheric ozone is primarily formed via reactions between

60

nitrogen oxides (NOx) and volatile organic compounds (VOCs). Reductions in dispersion could

61

increase both VOC and NOx concentrations, though impacts on ozone could be counterintuitive

62

due to non-linearities in ozone chemistry. Lowering air temperature decreases biogenic VOC

63

emissions from vegetation, potentially reducing ozone concentrations in urban areas where VOC

64

availability limits ozone formation24. Air temperature reduction also slows reactions that produce

65

ozone. Therefore, ozone concentrations are expected to decrease with lower temperatures25. 3 ACS Paragon Plus Environment


66

Secondary particulate matter includes sulfate, nitrate, ammonium, and secondary organic aerosols

67

(SOA). While temperature-dependent reactions that form secondary particulate matter should be

68

slower due to reduced temperatures, gas-particle partitioning for semi-volatile species (ammonium

69

nitrate and semi-volatile SOA) favors the particle phase26,27. The competing physical and chemical

70

processes lead to uncertainties in changes to air pollution concentrations induced by heat island

71

mitigation strategies.

72

The complexity of the aforementioned processes requires the use of sophisticated models that

73

resolve atmospheric physics and chemistry to predict how cool surface adoption would influence

74

city-level air quality. Using photochemical models, Taha et al.13,28 estimated that increasing city

75

surface albedo would effectively reduce ozone concentrations in Southern California and Central

76

California. Epstein et al.23 predicted that 8-hour daily maximum ozone concentrations would

77

decrease if cool roofs do not reflect more solar ultraviolet (UV) than do dark roofs; if solar UV

78

reflection is increased, ozone concentrations could rise.

79

Despite previous literature on the influence of cool roofs on ozone concentrations, there is only

80

one study that has investigated the influence of cool roofs on particulate matter23. They found that

81

increasing roof albedo would increase the annual mean concentrations of PM2.5, because reduced

82

ventilation would suppress dispersion of pollutants. However, they did not investigate (1) the

83

various physicochemical processes driving cool roof impacts on PM2.5 concentrations or (2) the

84

varying responses of different PM species (e.g., nitrate, sulfate, and organics) to cool roof adoption.

85

Cool walls are less studied than cool roofs. Zhang et al.29, for the first time, estimated the influence

86

of cool walls on urban climate, and systematically compared the effects of cool walls to cool roofs.

87

They found that adopting cool walls in Los Angeles would lead to daily average canyon air


Page 4 of 36

Page 5 of 36


88

temperature reductions of up to 0.40 K, which is slightly lower than that induced by adopting cool

89

roofs (0.43 K). However, the influence of cool walls on air quality has never been studied.

90

To address the aforementioned science knowledge gaps and inform policymaking on heat

91

mitigation strategies, we seek to (1) quantify and systematically compare the air quality effects of

92

adopting cool walls and roofs, and (2) investigate the physicochemical processes leading to

93

changes in particulate matter concentrations.

94

95

2 Method

96

2.1 Model description

97

We use the Weather Research and Forecasting model coupled with Chemistry Version 3.7 (WRF-

98

Chem V3.7), a state-of-the-science climate and air quality model, to estimate the impacts of

99

employing cool walls and roofs on air quality30. WRF-Chem has been widely used to study air

100

pollution in Southern California11,31,32. Table S2 summarizes our model configuration. The

101

following schemes are chosen for WRF physics: the Rapid Radiative Transfer Model (RRTM) for

102

long-wave radiation33, the Goddard shortwave radiation scheme34, the Lin et al. scheme35 for cloud

103

microphysics, the Grell 3D ensemble cumulus cloud scheme36, and the Yonsei University scheme

104

for the planetary boundary layer37.

105

Impervious fraction (Figure S3b) and land use classification in urban grid cells (Figure S3c) are

106

obtained from the National Land Cover Database (NLCD, 2006)38,39. The Noah land surface

107

model40 simulates land-atmosphere interactions in non-urban grid cells and for the pervious

108

portion of urban grid cells. The single-layer urban canopy model resolves urban physics and 5 ACS Paragon Plus Environment


109

simulates land-atmosphere interactions for the impervious portion of urban grid cells41. Urban grid

110

cells are classified as low-intensity residential (“Developed, Open Spaces” and “Developed, Low

111

Intensity” in NLCD), high-intensity residential (“Developed, Medium Intensity” in NLCD), and

112

commercial/industrial (“Developed, High Intensity” in NLCD). Urban morphology (i.e., roof

113

width, canyon floor width, and building height) is determined for each urban land use type based

114

on real-world building and street datasets for Los Angeles County, following Zhang et al29; the

115

datasets include National Urban Database and Access Portal (NUDAPT)42, the Los Angeles

116

Region Imagery Acquisition Consortium (LARIAC)43, and LA County Street and Address File44.

117

Since the default WRF-Chem is not compatible with the NLCD land use classification system, we

118

modify the model code to allow for use of NLCD urban land use types, following Fallmann et

119

al.45. We also implement satellite-based green vegetation fraction into the model following

120

Vahmani and Ban-Weiss21.

121

Gas phase chemistry is simulated using the Regional Atmospheric Chemistry Mechanism

122

(RACM)46 scheme, further updated by National Oceanic and Atmospheric Administration

123

(NOAA) Earth System Research Laboratory (ESRL)32. The RACM-ESRL scheme covers organic

124

and inorganic chemistry simulating 23 photolysis and 221 other chemical reactions47. The Modal

125

Aerosol Dynamics Model for Europe (MADE) simulates aerosol chemistry48. The volatility basis

126

set (VBS) is used for simulating secondary organic aerosols49.

127

We evaluate modeled ozone and PM2.5 concentrations against observations (Figures S1 and S2,

128

Table S1) from the Environmental Protection Agency’s Air Quality System in Section S1 of the

129

Supporting Information. Although our model underestimates ozone and PM2.5 concentrations at

130

higher concentrations, the bias in baseline concentrations does not necessarily lead to bias in

131

estimated changes induced by adopting cool surfaces.


Page 6 of 36

Page 7 of 36


132 133

2.2 Simulation domains

134

We simulate three nested domains (d1, d2, and d3, as shown in Figure S3a) with 30 layers in the

135

vertical at horizontal resolutions of 18 km, 6 km, and 2 km, respectively. The three domains each

136

cover the Southwestern United States (d1); Central and Southern California (d2); and Southern

137

California, including Los Angeles and San Diego (d3). Each outer domain provides boundary

138

conditions for the adjacent inner domain. In this paper we report results for the innermost domain.

139 140

2.3 Emission inventories

141

WRF-Chem requires gridded emissions inputs for each simulation. We use state-of-the-science

142

emission inventories from the South Coast Air Quality Management District (SCAQMD) and

143

California Air Resources Board (CARB) for the year 2012 (i.e., the most up-to-date inventories as

144

of writing this paper). For the outer two domains (d1 and d2), hourly emissions for the entire year

145

at 4-km resolution are provided by CARB for California50. Emissions outside California, but

146

within the simulation domain, are from National Emissions Inventory (NEI) by the Environmental

147

Protection Agency for the year 201151. For the innermost domain (d3), we use hourly emissions

148

for the entire year at 4-km resolution provided by SCAQMD52 . These emissions represent all

149

anthropogenic sources including motor vehicles; point sources such as refineries; and off-road

150

sources, such as construction. Emission inventories are regridded to match the grid for the modeled

151

domains and chemical speciation for RACM-ESRL and MADE/VBS mechanisms used in this

152

study. The Model of Emissions of Gases and Aerosols from Nature (MEGAN) is used to generate

153

temperature-dependent biogenic organic emissions53. Note that anthropogenic emissions are not 7 ACS Paragon Plus Environment


154

sensitive to ambient temperatures in our study. Though some anthropogenic emissions may be

155

temperature dependent (e.g., evaporative emissions of VOCs from gasoline powered vehicles),

156

this effect is not simulated in this study, as anthropogenic emissions are obtained directly from

157

input datasets.

158 159

2.4 Simulation design

160

To investigate the air quality effects of cool walls and roofs in Southern California, we simulate

161

three scenarios: CONTROL, where wall, roof, and pavement albedos are each set to 0.10;

162

COOL_WALL, where wall albedo is increased to 0.90; and COOL_ROOF, where roof albedo is

163

increased to 0.90. These cool surface albedos are intentionally chosen to quantify the upper bound

164

effects of adopting cool surfaces (i.e., increasing surface albedo by 0.80). Note that cool surface

165

albedos of actual cool walls and roofs are usually lower than 0.90. For example, the albedo of a

166

bright-white cool roof may decrease to 0.60–0.70 from an initial albedo of 0.80–0.90 after several

167

years of soiling and weathering54,55. In order to test the linearity of changes in air pollutant

168

concentrations to albedo increases, we add two additional scenarios where wall albedo and roof

169

albedo are each increased by 0.40.

170

Simulations are performed for 28 June 2012 to 11 July 2012, with the first five days discarded as

171

model “spin-up” to reduce the possible influence of inaccuracies in input initial conditions on our

172

analysis. We analyze the results from 00:00 local standard time (LST) on July 3 to 00:00 LST on

173

July 12. Section S3 in the Supporting Information demonstrates that the meteorology during our

174

analysis period is representative of summertime meteorology in Southern California. Thus, our

175

results are representative of changes induced by adopting cool surfaces under typical summertime 8 ACS Paragon Plus Environment

Page 8 of 36

Page 9 of 36


176

conditions in Southern California. The paired Student’s t-test (n = 9 analyzed days) is used to

177

assess whether the changes in cool surface scenarios relative to CONTROL are statistically

178

distinguishable from zero.

179 180 181

2.5 Method of attributing the changes in PM to ventilation versus other factors

182

Carbon monoxide (CO) is considered a chemically inert pollutant at urban scale, with

183

concentrations controlled by meteorological conditions. Therefore, past studies have used CO as

184

a tracer for transport and dispersion of pollutants9,56. Similarly, in our study, we use the increase

185

in CO concentration relative to CONTROL to quantify the increase in PM2.5 that is attributable to

186

ventilation (∆𝐶𝑃𝑀(vent)), as

∆𝐶𝑃𝑀(vent) =

∆𝐶CO 𝐶CO

× 𝐶𝑃𝑀

(1)

187

where ∆𝐶CO is the change in CO mixing ratio (ppbv) relative to CONTROL, 𝐶CO is the mixing

188

ratio (ppbv) of CO for CONTROL, 𝐶𝑃𝑀 is the concentration (μg m-3) of a PM species (i.e., total

189

PM2.5, sulfate, nitrate, elemental carbon, primary organic aerosol, anthropogenic secondary

190

organic aerosols, or biogenic secondary organic aerosols) for CONTROL, and all variables are

191

spatial averages over urban areas in Los Angeles County.

192

The change in concentration of a PM2.5 species that is not attributable to ventilation ∆𝐶𝑃𝑀(no vent)

193

(μg m-3) is then calculated as

∆𝐶𝑃𝑀(no vent) = ∆𝐶𝑃𝑀 ―

∆𝐶CO 𝐶CO

× 𝐶𝑃𝑀


(2)


194 195

In this way, we attribute increases in PM2.5 species to reductions in ventilation and changes in all

196

other processes. Note that while sea salt aerosols contribute to total PM2.5 concentrations, we omit

197

this species from the discussion because they are naturally produced and are not a public health

198

concern. Reductions in ventilation may also contribute to less vertical mixing and consequent

199

reductions in dry deposition of pollutants.

200

2.6 Caveats

201

In this study, we assume that adopting cool surfaces would not change reflectance in the UV

202

spectrum (280–400 nm). However, based on spectral reflectance measurements, UV reflectance

203

could increase from adopting cool roofs23. Increases in UV reflectance could enhance ozone

204

production and atmospheric oxidation capacity, which influences the formation of other secondary

205

pollutants. Therefore, changes in ozone are a result of competing effects among (a) ozone increases

206

induced by enhanced UV reflection, (b) ozone decreases induced by decreased temperatures, and

207

(c) ozone changes induced by reduced ventilation, which could affect the dispersion of ozone and

208

its precursors.

209

The influence of adopting cool surfaces is likely to vary by city due to differences in baseline

210

climate and land cover (e.g., vegetation distributions, building distributions, urban canyon

211

morphology). Also note that results might be different if simulated using another model or using

212

different parameterizations. For example, the single layer urban canopy model does not explicitly

213

resolve individual buildings.

214

Note that the urban morphology is derived using gross wall area (including windows) instead of

215

net wall area (excluding windows). In Los Angeles County, citywide ratio of net wall area to gross

216

wall area is 83%.29 In reality, windows may not be changed to cool colors. Therefore, a portion of 10 ACS Paragon Plus Environment

Page 10 of 36

Page 11 of 36


217

walls may not be able to be made solar reflective and our study may overestimate the influence of

218

adopting cool walls.

219

220

3 Results and discussion

221

By comparing changes in air pollutant concentrations for increasing albedo by 0.80 versus 0.40

222

relative to CONTROL, we find that the changes in air pollutant concentrations are approximately

223

linear to surface albedo change (Section S4 in the Supporting Information). Therefore, the results

224

reported for albedo increase of 0.80 can be interpolated to other albedo changes. For simplicity,

225

we report only results for COOL_WALL and COOL_ROOF in the main body.

226

3.1 Meteorological conditions

227

Figure 1 shows spatial distributions of near-surface air temperatures in the afternoon and evening.

228

(Diurnal cycles of near-surface air temperatures are shown in Figure S4.) For the CONTROL

229

scenario (Figure 1a), temperatures in inland areas are hotter than coastal areas, as expected.

230

Temperature reductions induced by adopting cool surfaces are higher in inland areas than in coastal

231

areas (Figure 1b,c). This is due to an accumulation effect in air temperature reduction as the sea

232

breeze advects air from the coast to inland.

233

Although total wall area in Los Angeles County is larger than roof area by a factor of 1.7, daily

234

average solar irradiance (W m-2) on walls is 38% of that on roofs29. In addition, 50-59% of the

235

solar radiation reflected by cool walls is absorbed by opposing walls or pavements, while all the

236

radiation reflected by cool roofs escapes the urban canopy in the model29. Therefore, daily average

237

temperature reductions induced by cool roofs (0.45 K) are larger than cool walls (0.24 K) over



238

urban areas in Los Angeles County, as shown in Table 1. Cool roofs are simulated to induce larger

239

temperature reductions than cool walls at both 14:00 LST (daytime) and 20:00 LST (nighttime).

240

Note that past studies investigating how air temperatures influence atmospheric chemistry often

241

report 2-meter air temperatures (“T2”)8,45. However, 2-meter air temperature is a diagnostic

242

variable that is not used in model calculations of atmospheric chemistry. The chemistry model

243

actually uses the four-dimensional (x, y, z, t) atmospheric temperature. Therefore, we present

244

temperatures in the lowest atmospheric layer as “near-surface air temperature” rather than “T2.”

245

Figures S5 and S6 show diurnal cycles and spatial maps of 10-meter horizontal wind speeds, and

246

Figure S7 shows horizontal wind vectors. For the CONTROL scenario, winds are southwesterly

247

from coast to inland and wind speed is higher during daytime than nighttime. As shown in Table

248

1, spatially averaged wind speed in urban areas is 4.2 m s-1 and 2.3 m s-1 at 14:00 LST and 20:00

249

LST, respectively. Simulations predict that adopting cool walls (roofs) decreases onshore wind

250

speeds by 0.06 (0.21) m s-1 at 14:00 LST and 0.08 (0.09) m s-1 at 20:00 LST. This can be explained

251

by the reduced temperature difference between urban land and ocean, which is a driver for the sea

252

breeze.

253

Figure S8 show the diurnal cycle of planetary boundary layer (PBL) height. PBL height reaches

254

its maximum at 12:00 LST. Adopting cool walls reduces PBL height by 3-7% at most times of

255

day. Adopting cool roofs reduces PBL height by about 5% at night and about 10% during the day.

256

The reduction in PBL height can be attributed to decreases in surface temperatures and consequent

257

reductions in convection. Decreases in wind speeds and PBL height tend to reduce ventilation for

258

pollutants. The influence of changes in ventilation on particulate matter is discussed in Section

259

3.5.1.


Page 12 of 36

Page 13 of 36


260 261

3.2 Spatial distribution of ozone concentrations

262

Figure 2 shows the spatial distribution of daily maximum 8-hour average (MDA8) ozone

263

concentrations. MDA8 ozone is regulated by the National Ambient Air Quality Standards of the

264

Environmental Protection Agency. For the CONTROL scenario, the ozone concentration over

265

urban areas is lower than rural areas because (a) southwesterly winds transport ozone and its

266

precursors from the coast to the inland areas, creating an accumulation effect as this secondary

267

pollutant is generated in the atmosphere; and (b) nitric oxide emissions in urban areas can titrate

268

ozone. Adopting cool walls can decrease the spatially averaged MDA8 ozone concentration by

269

0.35 ppbv in the urban areas of Los Angeles County (Table 1). These decreases in ozone

270

concentrations are likely due to reductions in temperature-dependent ozone formation. Adopting

271

cool roofs can lead to a greater reduction in MDA8 ozone concentration (0.83 ppbv) than cool

272

walls. This is likely because the near-surface air temperature reductions induced by cool roofs are

273

larger than that induced by cool walls during daytime (Figure S4) and thus the decreases in reaction

274

rates for ozone production are larger for COOL_ROOF than COOL_WALL relative to CONTROL.

275

As mentioned in Section 2.5, we assume that the UV reflectance of cool surfaces is the same as

276

dark surfaces. Similarly, Epstein et al23 report reductions in ozone concentrations in most Southern

277

California regions due to adopting cool roofs when UV reflectance is assumed to be held constant.

278

(Note that they also find that ozone concentrations could increase if the difference in UV

279

reflectance between cool and dark roofs follows an upper bound scenario.)

280



281

3.3 Spatial distribution of PM2.5 species

282

Figure 3 shows the spatial distribution of daily average PM2.5 species concentrations and changes

283

due to adopting cool surfaces. PM2.5 concentrations reported here represent dry particle mass.

284

Spatial distributions of PM2.5 species concentrations in the CONTROL scenario are mainly

285

attributable to spatial patterns in emissions and meteorology. For example, when the sea breeze

286

advects air from the coast to inland, EC, a primary pollutant, accumulates, leading to higher

287

concentrations in locations further east. For spatial distributions of sulfate concentrations, there

288

are higher concentrations near the ports of Los Angeles and Long Beach that are likely due to hot

289

spots in SO2 (the precursor of secondary sulfate) and primary sulfate emissions from ships and

290

power plants (Figure S9). Meanwhile, southwesterly winds then transport these emissions to

291

downtown Los Angeles, making concentrations downtown greater than those further east. The

292

spatial variability of anthropogenic and biogenic SOA is relatively small compared to other species.

293

The concentrations of total PM2.5 and each individual species increase due to cool surface adoption

294

(Figure 3). The increase in each PM2.5 species induced by adopting cool roofs is larger than that

295

induced by cool walls, though their spatial patterns are similar. Spatial distributions of increases

296

in total PM2.5 and individual species (except nitrate) are consistent with the spatial patterns of

297

absolute concentrations in the CONTROL scenario. In other words, the regions with the highest

298

baseline concentrations show the largest changes in PM2.5 due to meteorological shifts from cool

299

surface adoption. The exception is for nitrate, which shows larger increases in urban residential

300

areas northeast of downtown where baseline concentrations are low, rather than downtown where

301

baseline concentrations are the highest in CONTROL. This is likely due to the greater temperature

302

reductions in regions northeast of downtown Los Angeles relative to downtown, especially at night

303

(Figure 1b,c). The processes leading to nitrate increases will be discussed in Section 3.5. The 14 ACS Paragon Plus Environment

Page 14 of 36

Page 15 of 36


304

increase in SOA is relatively smaller than other species, which will also be explained in Section

305

3.5.

306

3.4 Diurnal cycles of PM2.5 species concentrations

307

Figure 4 shows the diurnal cycles of spatially averaged PM2.5 species concentrations and their

308

changes in the urban areas of Los Angeles County. For the CONTROL scenario, PM2.5, nitrate,

309

ammonium, sulfate, EC, and primary organic aerosol (POA) concentrations reach their maximum

310

between 08:00 and 09:00 LST and their minimum at 16:00 LST, while biogenic and anthropogenic

311

SOA reach their maximum near 14:00 LST and their minimum at night. The diurnal cycles of

312

PM2.5 concentrations can be attributed to the diurnal variation of (1) emissions (Figure S10); (2)

313

PBL height (Figure S8a), which peaks at 12:00 LST; (3) wind speed (Figure S5), which peaks at

314

14:00 LST; and (4) photochemical reaction rates for secondary species that depend on UV

315

radiation and temperature (Figure S4a).

316

Raising roof or wall albedo leads to increases in concentrations of total PM2.5 and most individual

317

species (except for biogenic SOA) throughout the day (Figure 4). Increases in nitrate

318

concentrations are the largest among all PM2.5 species, followed by increases in POA, sulfate, and

319

ammonium, while the increases in concentrations of other PM2.5 species are relatively small. The

320

changes in speciated PM2.5 concentrations due to adopting cool walls or roofs vary by time of day,

321

and the mechanisms contributing to the changes will be discussed in Section 3.5. For all PM2.5

322

species except biogenic SOA, increases in PM2.5 concentrations induced by adopting cool roofs

323

are larger than those induced by adopting cool walls during most daytime hours (07:00-19:00 LST).

324

On daily average, cool roof adoption contributes to greater increases in particulate matter than cool

325

wall adoption (Table 1) for total PM2.5 and each species. Daily average increases in total PM2.5

326

concentrations are simulated to be 0.62 (0.85) μg m-3 upon increasing wall (roof) albedo by 0.80 15 ACS Paragon Plus Environment


327

in July in Los Angeles County. Compared to the national annual and 24-hour PM2.5 standard of 12

328

μg m−3 and 35 μg m−3, respectively, increases in PM2.5 concentrations reported here have the

329

potential for increasing exceedance days of federal air quality standards. The grid cell containing

330

Mira Loma (i.e., the most polluted PM2.5 monitoring station in Southern California) is simulated

331

to have PM2.5 increases of 0.84 (1.05) μg m−3 due to adopting cool walls (roofs) in summer. Epstein

332

et al. (2018) estimate that annual average PM2.5 concentrations at Mira Loma would increase by

333

0.19 μg m-3 due to adopting cool roofs, which they compute would result in an increase of 2/3

334

exceedance day for the 24-hr federal PM2.5 standard. (The number of exceedance days is not an

335

integer because they report 3x3 cell moving averages.) Thus, even though these changes may look

336

small, they have the potential to increase the annual number of days exceeding air quality standards

337

and are therefore important for regulatory agencies in controlling PM2.5 pollution.

338

3.5 Mechanisms that lead to changes in PM2.5 concentrations

339

As mentioned in the introduction, adopting cool surfaces can influence PM2.5 concentrations

340

mainly via (1) reducing ventilation, (2) slowing temperature dependent reactions and emissions,

341

and (3) increasing the likelihood that semi-volatile species will partition to particle phase. In the

342

following sections we report on the relative importance of these pathways.

343

3.5.1 Ventilation

344

For primary pollutants such as elemental carbon (EC), mass concentrations depend highly on

345

ventilation and are insensitive to atmospheric chemistry in the model. (Note that strictly speaking,

346

hydrophilic species can coat EC and increase its hygroscopicity, enabling the in-cloud wet

347

scavenging of EC57. This so-called “aging process” depends on temperature-dependent

348

atmospheric photochemical reactions that form hydrophilic species, such as sulfate. However, the 16 ACS Paragon Plus Environment

Page 16 of 36

Page 17 of 36


349

aging of EC should not be a very important process during summer when there is little precipitation

350

in the Los Angeles Basin.) Decreases in ventilation (Section 3.1) impede the dilution and transport

351

of pollutants in source regions and may also reduce dry deposition, leading to increases in near-

352

surface pollutant concentrations. This ventilation effect is driven by vertical and horizontal mixing

353

of pollutants in the planetary boundary layer, which can be investigated using PBL height and

354

surface wind speeds, respectively. Figure 5 shows that fractional increase in EC is positively

355

correlated with the fractional reductions in PBL height and 10-meter wind speed. Fractional

356

reduction in PBL height can explain 42% of the variability in the fractional increase in EC

357

concentrations for both COOL_WALL – CONTROL and COOL_ROOF – CONTROL. Fractional

358

reduction in horizontal wind speed explains 17% (79%) of the variability in fractional increase of

359

EC concentrations due to adopting cool walls (roofs).

360

3.5.2 Quantifying the relative importance of ventilation versus other factors for driving

361

changes in PM

362

Following the method described in Section 2.5, we quantify increases in PM2.5 species that can be

363

attributed to reductions in ventilation and changes in other processes. As indicated in Figure 4,

364

after removing the effects of ventilation, the change in spatially averaged EC and POA is close to

365

zero. Therefore, increases in primary pollutant (EC and POA) concentrations are attributable to

366

suppressed ventilation. A large fraction of the increase in sulfate from cool surface adoption can

367

be attributed to suppressed ventilation. Other driving processes can affect sulfate concentrations:

368

(a) reductions in temperature-dependent reaction rates would decrease sulfate production; and (b)

369

changes in cloud cover can also influence in-cloud SO2 oxidation, which occurs faster than gas-

370

phase oxidation of SO2 if clouds are present. When the ventilation effect is excluded, sulfate

371

concentrations slightly increase from 04:00 to 14:00 LST but decrease at most other hours, due to 17 ACS Paragon Plus Environment


372

adopting cool surfaces. Nevertheless, ventilation is the dominant process leading to sulfate

373

increases, contributing to 76 % (91%) of the daily average increase for COOL_WALL –

374

CONTROL (COOL_ROOF – CONTROL).

375

On the other hand, the ventilation effect accounts for a small portion of the increase in semi-volatile

376

species such as nitrate and ammonium (in the form of ammonium nitrate). Concentrations of these

377

particulate species rise drastically even when the ventilation effect is excluded. This is because the

378

reaction between gas-phase ammonia and nitric acid that forms particulate nitrate is reversible,

379

and the equilibrium constant for the reaction is highly temperature dependent. Temperature

380

reductions would cause gas to particle conversion and increase the concentrations of ammonium

381

nitrate26. Note that the amount of nitrate at equilibrium has a non-linear relationship with

382

temperature. Thus, the relationship between increase in nitrate concentration due to gas-to-particle

383

conversion (Figure 4) and temperature reduction is not linear; the increase in nitrate depends not

384

only on the magnitude of temperature reduction but also the baseline temperature. In contrast to

385

shifting equilibrium of the reaction between nitric acid and ammonia, which would increase nitrate,

386

cool surfaces adoption may also reduce photochemistry and impede the formation of nitric acid

387

precursors (i.e., OH and NO2) during the day, leading to a reduction in nitrate. Increased gas-to-

388

particle conversion and suppressed ventilation outweigh reductions in photochemistry, leading to

389

overall increases in nitrate concentrations (Figure 4).

390

For secondary organic aerosols (SOA), reductions in ventilation should lead to increases in SOA,

391

while temperature decreases would be expected to cause (a) increases in gas-to-particle conversion

392

for semi-volatile species, which would lead to SOA increases, and (b) reduced rates of

393

temperature-dependent reactions, which would lead to SOA decreases. Biogenic SOA may also

394

be influenced by reductions in temperature dependent VOC emissions (e.g., isoprene) from 18 ACS Paragon Plus Environment

Page 18 of 36

Page 19 of 36


395

vegetation (Figure S11). As shown in Figure 4, both anthropogenic and biogenic SOA increase

396

when including the influence of changes in ventilation, but decrease when ventilation changes are

397

excluded. Daily average SOA concentrations increase by 0.018 (0.046) μg m-3 for COOL_WALL

398

(COOL_ROOF) relative to CONTROL. After removing the ventilation effect, daily average SOA

399

concentrations decrease by 0.057 (0.071) μg m-3 for COOL_WALL (COOL_ROOF) relative to

400

CONTROL. This means that SOA reductions induced by slowed temperature dependent reactions

401

and biogenic emissions outweigh the expected increases in semi-volatile SOA species due to phase

402

partitioning. On the other hand, increases in SOA due to suppressed ventilation and increased gas-

403

to-particle conversion outweigh decreases in SOA due to reduced reaction and emission rates.

404

These competing effects lead to an overall increase in SOA concentrations, although fractional

405

increases are small relative to other species.

406

In this paper, we discuss the climate and air quality implications of cool roofs and cool walls,

407

which have been used in cities to reduce temperatures and thus combat global warming and urban

408

heat islands. Our results show that reductions in urban surface temperatures lead to both co-

409

benefits of reduced ozone concentrations and penalties of increased PM2.5 concentrations,

410

potentially changing the exceedance days of federal air quality standard in the Los Angeles Basin.

411

We suggest further studies to assess the air quality effects of other heat strategies and the effects

412

in other cities. For policy makers, it is important to assess the effects of environmental solutions

413

from a systematic perspective, i.e., looking at heat mitigation impacts not just from a climate

414

perspective but also from an air quality perspective.

415



416

Acknowledgements

417

This research was supported by the California Energy Commission under contract EPC-14-010,

418

and the National Science Foundation under grants CBET-1512429 and 1752522. This work was

419

also supported by the Assistant Secretary for Energy Efficiency and Renewable Energy, Building

420

Technologies Office of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.

421

Computation for the work described in this paper was supported by the University of Southern

422

California’s Center for High-Performance Computing (https://hpcc.usc.edu/). We thank Scott

423

Epstein and Sang-Mi Lee at South Coast Air Quality Management District for providing us

424

emission datasets. We also thank Pouya Vahmani, Haley Gilbert, Pablo Rosado, Hugo Destaillats,

425

and Xiaochen Tang at Lawrence Berkeley National Laboratory, Dan Li at Boston University,

426

Ravan Ahmadov and Stu McKeen at National Oceanic and Atmospheric Administration, Joachim

427

Fallmann at Johannes Gutenberg University Mainz, Gert-Jan Steeneveld at Wageningen

428

University, and Jan Kleissl at University of California, San Diego for their helpful suggestions.

429

Supporting Information

430

Model evaluation, description of the simulated period, and further detail on results. This material

431

is available free of charge via the Internet at http://pubs.acs.org.

432

433

References

434

(1)

Seto, K. C.; Guneralp, B.; Hutyra, L. R. Global Forecasts of Urban Expansion to 2030 and Direct


Page 20 of 36

Page 21 of 36


435 436

Impacts on Biodiversity and Carbon Pools. Proc. Natl. Acad. Sci. 2012, 109 (40), 16083–16088. (2)

437 438

Oke, T. R. City Size and the Urban Heat Island. Atmos. Environ. Pergamon Pres 1973, 7, 769– 779.

(3)

Civerolo, K.; Hogrefe, C.; Lynn, B.; Rosenthal, J.; Ku, J. Y.; Solecki, W.; Cox, J.; Small, C.;

439

Rosenzweig, C.; Goldberg, R.; Knowlton, K.; Kinney, P. Estimating the Effects of Increased

440

Urbanization on Surface Meteorology and Ozone Concentrations in the New York City

441

Metropolitan Region. Atmos. Environ. 2007, 41 (9), 1803–1818.

442

(4)

443 444

Li, J.; Georgescu, M.; Hyde, P.; Mahalov, A.; Moustaoui, M. Achieving Accurate Simulations of Urban Impacts on Ozone at High Resolution. Environ. Res. Lett. 2014, 9 (11), 114019.

(5)

Wang, X. M.; Lin, W. S.; Yang, L. M.; Deng, R. R.; Lin, H. A Numerical Study of Influences of

445

Urban Land-Use Change on Ozone Distribution over the Pearl River Delta Region, China. Tellus

446

B Chem. Phys. Meteorol. 2007, 59 (3), 633–641.

447

(6)

448 449

688–698. (7)

450 451

Stone, B. Urban Sprawl and Air Quality in Large US Cities. J. Environ. Manage. 2008, 86 (4),

Zhong, S.; Qian, Y.; Sarangi, C.; Zhao, C.; Leung, R.; Wang, H. Urbanization Effect on Winter Haze in the Yangtze River Delta Region of China. Geophys. Res. Lett. 2018, 45 (13), 6710–6718.

(8)

Tao, W.; Liu, J.; Ban-Weiss, G. A.; Hauglustaine, D. A.; Zhang, L.; Zhang, Q.; Cheng, Y.; Yu, Y.;

452

Tao, S. Effects of Urban Land Expansion on the Regional Meteorology and Air Quality of Eastern

453

China. Atmos. Chem. Phys. 2015, 15 (15), 8597–8614.

454

(9)

Tao, W.; Liu, J.; Ban-Weiss, G. A.; Zhang, L.; Zhang, J.; Yi, K.; Tao, S. Potential Impacts of

455

Urban Land Expansion on Asian Airborne Pollutant Outflows. J. Geophys. Res. 2017, 122 (14),

456

7646–7663.

457

(10)

Sarrat, C.; Lemonsu, A.; Masson, V.; Guedalia, D. Impact of Urban Heat Island on Regional 21 ACS Paragon Plus Environment


458 459

Atmospheric Pollution. Atmos. Environ. 2006, 40 (10), 1743–1758. (11)

460 461

and Air Quality in Southern California. Atmos. Chem. Phys. 2019, 19 (7), 4439–4457. (12)

462 463

(13)

(14)

(15)

Taha, H. Meteorological, Air-Quality, and Emission-Equivalence Impacts of Urban Heat Island Control in California. Sustain. Cities Soc. 2015, 19, 207–221.

(16)

470 471

Taha, H.; Chang, S.; Akbari, H. Meteorological and Air Quality Impacts of Heat Island Mitigation Measures in Three U.S. Cities. Energy 2000, No. April.

468 469

Taha, H. Meso-Urban Meteorological and Photochemical Modeling of Heat Island Mitigation. Atmos. Environ. 2008, 42 (38), 8795–8809.

466 467

Rosenfeld, A. H.; Akbari, H.; Romm, J. J.; Pomerantz, M. Cool Communities: Strategies for Heat Island Mitigation and Smog Reduction. Energy Build. 1998, 28 (1), 51–62.

464 465

Li, Y.; Zhang, J.; Sailor, D. J.; Ban-weiss, G. A. Effects of Urbanization on Regional Meteorology

Zhang, J.; Zhang, K.; Liu, J.; Ban-Weiss, G. Revisiting the Climate Impacts of Cool Roofs around the Globe Using an Earth System Model. Environ. Res. Lett. 2016, 11 (8), 084014.

(17)

Mohegh, A.; Rosado, P.; Jin, L.; Millstein, D.; Levinson, R.; Ban-Weiss, G. Modeling the Climate

472

Impacts of Deploying Solar Reflective Cool Pavements in California Cities. J. Geophys. Res.

473

2017, 122 (13), 6798–6817.

474

(18)

Taha, H.; Konopacki, S.; Gabersek, S. Impacts of Large-Scale Surface Modifications on

475

Meteorological Conditions and Energy Use: A 10-Region Modeling Study. Theor. Appl. Climatol.

476

1999, 62 (3–4), 175–185.

477

(19)

Georgescu, M.; Morefield, P. E.; Bierwagen, B. G.; Weaver, C. P. Urban Adaptation Can Roll

478

Back Warming of Emerging Megapolitan Regions. Proc. Natl. Acad. Sci. 2014, 111 (8), 2909–

479

2914.


Page 22 of 36

Page 23 of 36

480


(20)

481 482

Li, D.; Bou-Zeid, E.; Oppenheimer, M. The Effectiveness of Cool and Green Roofs as Urban Heat Island Mitigation Strategies. Environ. Res. Lett. 2014, 9 (5), 055002.

(21)

Vahmani, P.; Ban-Weiss, G. Climatic Consequences of Adopting Drought-Tolerant Vegetation

483

over Los Angeles as a Response to California Drought. Geophys. Res. Lett. 2016, 43 (15), 8240–

484

8249.

485

(22)

Mohegh, A.; Levinson, R.; Taha, H.; Gilbert, H.; Zhang, J.; Li, Y.; Tang, T.; Ban-Weiss, G.

486

Observational Evidence of Neighborhood Scale Reductions in Air Temperature Associated with

487

Increases in Roof Albedo. Climate 2018, 6 (4), 98.

488

(23)

Epstein, S. A.; Lee, S.-M.; Katzenstein, A. S.; Carreras-Sospedra, M.; Zhang, X.; Farina, S. C.;

489

Vahmani, P.; Fine, P. M.; Ban-Weiss, G. Air-Quality Implications of Widespread Adoption of

490

Cool Roofs on Ozone and Particulate Matter in Southern California. Proc. Natl. Acad. Sci. 2017,

491

201703560.

492

(24)

493 494

Effects of Urban Vegetation on Air Pollution. Air Pollut. Model. its Appl. XII 1998, 399–407. (25)

495 496

Nowak, D. J.; McHale, P. J.; Ibarra, M.; Crane, D.; Stevens, J. C.; Luley, C. J. Modeling the

Taha, H. Modeling the Impacts of Large-Scale Albedo Changes on Ozone Air Quality in the South Coast Air Basin. Atmos. Environ. 1997, 31 (11), 1667–1676.

(26)

Moya, M.; Ansari, A. S.; Pandis, S. N. Partitioning of Nitrate and Ammonium between the Gas

497

and Particulate Phases during the 1997 IMADA-AVER Study in Mexico City. Atmos. Environ.

498

2001, 35 (10), 1791–1804.

499

(27)

Pun, B. K.; Griffin, R. J.; Seigneur, C.; Seinfeld, J. H. Secondary Organic Aerosol 2.

500

Thermodynamic Model for Gas/Particle Partitioning of Molecular Constituents. J. Geophys. Res.

501

Atmos. 2002, 107 (D17), AAC-4.

502

(28)

Taha, H. Urban Surface Modification as a Potential Ozone Air-Quality Improvement Strategy in 23 ACS Paragon Plus Environment


503 504

California: A Mesoscale Modelling Study. Boundary-Layer Meteorol. 2008, 127 (2), 219–239. (29)

Zhang, J.; Mohegh, A.; Li, Y.; Levinson, R.; Ban-Weiss, G. Systematic Comparison of the

505

Influence of Cool Wall versus Cool Roof Adoption on Urban Climate in the Los Angeles Basin.

506

Environ. Sci. Technol. 2018, 52 (19), 11188–11197.

507

(30)

Grell, G. A.; Peckham, S. E.; Schmitz, R.; McKeen, S. A.; Frost, G.; Skamarock, W. C.; Eder, B.

508

Fully Coupled “Online” Chemistry within the WRF Model. Atmos. Environ. 2005, 39 (37), 6957–

509

6975.

510

(31)

Chen, D.; Li, Q.; Stutz, J.; Mao, Y.; Zhang, L.; Pikelnaya, O.; Tsai, J. Y.; Haman, C.; Lefer, B.;

511

Rappenglück, B.; Alvarez, S. L.; Neuman, J. A.; Flynn, J.; Roberts, J. M.; Nowak, J. B.; de Gouw,

512

J.; Holloway, J.; Wagner, N. L.; Veres, P.; Brown, S. S.; Ryerson, T. B.; Warneke, C.; Pollack, I.

513

B. WRF-Chem Simulation of NOx and O3 in the L.A. Basin during CalNex-2010. Atmos.

514

Environ. 2013, 81, 421–432.

515

(32)

Kim, S. W.; Heckel, A.; Frost, G. J.; Richter, A.; Gleason, J.; Burrows, J. P.; McKeen, S.; Hsie, E.

516

Y.; Granier, C.; Trainer, M. NO2 Columns in the Western United States Observed from Space and

517

Simulated by a Regional Chemistry Model and Their Implications for NOx Emissions. J. Geophys.

518

Res. Atmos. 2009, 114 (11), 1–29.

519

(33)

Mlawer, E. J.; Taubman, S. J.; Brown, P. D.; Iacono, M. J.; Clough, S. A. Radiative Transfer for

520

Inhomogeneous Atmospheres: RRTM, a Validated Correlated-k Model for the Longwave. J.

521

Geophys. Res. Atmos. 1997, 102 (D14), 16663–16682.

522

(34)

523 524 525

Chou, M.; Suarez, M. J. A Solar Radiation Parameterization for Atmospheric Studies. Technical Report Series on Global Modeling and Data Assimilation 1999.

(35)

Lin, Y.-L.; Farley, R. D.; Orville, H. D.; Lin, Y.-L.; Farley, R. D.; Orville, H. D. Bulk Parameterization of the Snow Field in a Cloud Model. J. Clim. Appl. Meteorol. 1983, 22 (6),


Page 24 of 36

Page 25 of 36


526 527

1065–1092. (36)

528 529

Ensemble and Data Assimilation Techniques. Geophys. Res. Lett. 2002, 29 (14), 38-1-38–4. (37)

530 531

Grell, G. A.; Dévényi, D. A Generalized Approach to Parameterizing Convection Combining

Hong, S.-Y.; Noh, Y.; Dudhia, J. A New Vertical Diffusion Package with an Explicit Treatment of Entrainment Processes. Mon. Weather Rev. 2006, 134 (9), 2318–2341.

(38)

Wickham, J. D.; Stehman, S. V.; Gass, L.; Dewitz, J.; Fry, J. A.; Wade, T. G. Accuracy

532

Assessment of NLCD 2006 Land Cover and Impervious Surface. Remote Sens. Environ. 2013,

533

130, 294–304.

534

(39)

Fry, J. A.; Xian, G.; Jin, S.; Dewitz, J. A.; Homer, C. G.; Yang, L.; Barnes, C. A.; Herold, N. D.;

535

Wickham, J. D. Completion of the 2006 National Land Cover Database for the Conterminous

536

United States. Photogramm. Eng. Remote Sensing 2011, 77, 858–566.

537

(40)

Chen, F.; Dudhia, J.; Chen, F.; Dudhia, J. Coupling an Advanced Land Surface–Hydrology Model

538

with the Penn State–NCAR MM5 Modeling System. Part I: Model Implementation and

539

Sensitivity. Mon. Weather Rev. 2001, 129 (4), 569–585.

540

(41)

Kusaka, H.; Kondo, H.; Kikegawa, Y.; Kimura, F. A Simple Single-Layer Urban Canopy Model

541

for Atmospheric Models: Comparison with Multi-Layer and Slab Models. Boundary-Layer

542

Meteorol. 2001, 101 (3), 329–358.

543

(42)

Ching, J.; Brown, M.; Burian, S.; Chen, F.; Cionco, R.; Hanna, A.; Hultgren, T.; McPherson, T.;

544

Sailor, D.; Taha, H.; Williams, D. National Urban Database and Access Portal Tool. Bull. Am.

545

Meteorol. Soc. 2009, 90 (8), 1157–1168.

546

(43)

547 548

LARIAC. Countywide Building Outlines https://egis3.lacounty.gov/dataportal/lariac/ (accessed Jul 11, 2017).

(44)

CAMS. LA County Street & Address File https://egis3.lacounty.gov/dataportal/2014/06/16/201125 ACS Paragon Plus Environment


549 550

la-county-street-centerline-street-address-file (accessed Aug 3, 2017). (45)

551 552

on Air Quality. Atmos. Environ. 2016, 125, 199–211. (46)

553 554

Fallmann, J.; Forkel, R.; Emeis, S. Secondary Effects of Urban Heat Island Mitigation Measures

Stockwell, W. R.; Kirchner, F.; Kuhn, M.; Seefeld, S. A New Mechanism for Regional Atmospheric Chemistry Modeling. J. Geophys. Res. Atmos. 1997, 102 (D22), 25847–25879.

(47)

Ahmadov, R.; McKeen, S.; Trainer, M.; Banta, R.; Brewer, A.; Brown, S.; Edwards, P. M.; De

555

Gouw, J. A.; Frost, G. J.; Gilman, J.; Helmig, D.; Johnson, B.; Karion, A.; Koss, A.; Langford, A.;

556

Lerner, B.; Olson, J.; Oltmans, S.; Peischl, J.; Pétron, G.; Pichugina, Y.; Roberts, J. M.; Ryerson,

557

T.; Schnell, R.; Senff, C.; Sweeney, C.; Thompson, C.; Veres, P. R.; Warneke, C.; Wild, R.;

558

Williams, E. J.; Yuan, B.; Zamora, R. Understanding High Wintertime Ozone Pollution Events in

559

an Oil- and Natural Gas-Producing Region of the Western US. Atmos. Chem. Phys. 2015, 15 (1),

560

411–429.

561

(48)

Ackermann, I. J.; Hass, H.; Memmesheimer, M.; Ebel, A.; Binkowski, F. S.; Shankar, U. Modal

562

Aerosol Dynamics Model for Europe Development and First Applications. Atmos. Environ. 1998,

563

32 (17), 2981–2999.

564

(49)

Ahmadov, R.; McKeen, S. A.; Robinson, A. L.; Bahreini, R.; Middlebrook, A. M.; de Gouw, J. A.;

565

Meagher, J.; Hsie, E.-Y.; Edgerton, E.; Shaw, S.; Trainer, M. A Volatility Basis Set Model for

566

Summertime Secondary Organic Aerosols over the Eastern United States in 2006. J. Geophys.

567

Res. Atmos. 2012, 117 (D6), n/a-n/a.

568

(50)

569 570 571

California Air Resources Board. ARB’s Emission Inventory Activities https://www.arb.ca.gov/ei/ei.htm (accessed Jun 18, 2017).

(51)

USEPA. Profile of the 2011 National Air Emissions Inventory (U.S. EPA 2011 NEI Version 1.0). 2014, No. April, 23.


Page 26 of 36

Page 27 of 36

572


(52)

South Coast Air Quality Management District. Final 2016 Air Quality Management Plan,

573

Appendix III: Base and Future Year Emission Inventory http://www.aqmd.gov/docs/default-

574

source/clean-air-plans/air-quality-management-plans/2016-air-quality-management-plan/final-

575

2016-aqmp/appendix-iii.pdf (accessed Jun 18, 2018).

576

(53)

Guenther, A.; Karl, T.; Harley, P.; Wiedinmyer, C.; Palmer, P. I.; Geron, C. Estimates of Global

577

Terrestrial Isoprene Emissions Using MEGAN (Model of Emissions of Gases and Aerosols from

578

Nature). Atmos. Chem. Phys. 2006, 6 (11), 3181–3210.

579

(54)

Sleiman, M.; Ban-Weiss, G.; Gilbert, H. E.; François, D.; Berdahl, P.; Kirchstetter, T. W.;

580

Destaillats, H.; Levinson, R. Soiling of Building Envelope Surfaces and Its Effect on Solar

581

Reflectance - Part I: Analysis of Roofing Product Databases. Sol. Energy Mater. Sol. Cells 2011,

582

95 (12), 3385–3399.

583

(55)

584 585

Berdahl, P.; Akbari, H.; Levinson, R.; Miller, W. A. Weathering of Roofing Materials - An Overview. Constr. Build. Mater. 2008, 22 (4), 423–433.

(56)

Zhang, Q.; Quan, J.; Tie, X.; Li, X.; Liu, Q.; Gao, Y.; Zhao, D. Effects of Meteorology and

586

Secondary Particle Formation on Visibility during Heavy Haze Events in Beijing, China. Sci.

587

Total Environ. 2015, 502, 578–584.

588 589

(57)

Zhang, J.; Liu, J.; Tao, S.; Ban-Weiss, G. A. Long-Range Transport of Black Carbon to the Pacific Ocean and Its Dependence on Aging Timescale. Atmos. Chem. Phys. 2015, 15 (20), 11521–11535.

590 591



Page 28 of 36

592 593 594 595 596 597

Table 1. Spatially averaged meteorological variables and pollutant concentrations for the CONTROL scenario, and the change relative to CONTROL for COOL_WALL and COOL_ROOF. Values represent spatial averages in Los Angeles County (shown in Figure S3b) for urban grid cells from 00:00 LST on July 3 to 00:00 LST on July 12. CONTROL

COOL_WALL minus CONTROL

COOL_ROOF minus CONTROL

Daily average near-surface air temperature a (K)

292.85

-0.24

-0.45

10-meter wind speed at 14:00 LST (m s-1)

4.15

-0.06

-0.21

10-meter wind speed at 20:00 LST (m s-1)

2.28

-0.08

-0.09

Daily maximum 8-hour average ozone concentration (ppbv) Daily average PM2.5 concentration (μg m-3)

38.47

-0.35

-0.83

12.25

0.62

0.85

Daily average nitrate concentration b (μg m-3)

0.89

0.11

0.18

Daily average ammonium concentration b (μg m-3)

0.98

0.07

0.10

Daily average sulfate concentration b (μg m-3)

1.91

0.11

0.13

Daily average EC concentration b (μg m-3)

0.87

0.05

0.06

Daily average anthropogenic SOA concentration b (μg m-3)

1.22

0.01

0.04

Daily average biogenic SOA concentration b (μg m-3)

0.51

0.01

0.01

Daily average POA concentration b (μg m-3)

1.90

0.12

0.14

598

a

599 600

b

Near-surface air temperature refers to the temperature in the lowest atmospheric layer.

Mass concentrations for particles with diameter less than 2.5 μm (i.e., nuclei and accumulation mode) are included for each species.

601


Page 29 of 36


602

603 604 605 606

Figure 1. Spatially resolved near-surface air temperatures (K) at 14:00 LST and 20:00 LST for (a) the CONTROL scenario, and the difference relative to CONTROL for (b) COOL_WALL and (c) COOL_ROOF. Values are temporally averaged over the period of 00:00 LST on July 3 to 00:00 LST on July 12.

607



609 610

611 612 613 614 615 616

Figure 2. Spatially resolved daily maximum 8-hour average (MDA8) ozone concentrations (ppbv) for (a) the CONTROL scenario, and changes relative to CONTROL for (b) COOL_WALL and (c) COOL_ROOF. Changes that are not statistically distinguishable from zero (see section 2.4 for details on statistical analysis) in (b) and (c) are dotted. Values are temporally averaged over the period of 00:00 LST on July 3 to 00:00 LST on July 12.

617


Page 30 of 36

Page 31 of 36


618



619 620 621 622 623

Figure 3. Daily average PM2.5 concentrations (μg m-3) by species for CONTROL (left column), as well as the differences for COOL_WALL – CONTROL (middle column) and COOL_ROOF – CONTROL (right column). Differences that are not statistically distinguishable from zero (see Section 2.4 for details on statistical analysis) are shaded in gray (middle and right columns of panels). Values are temporally averaged over the period of 00:00 LST on July 3 to 00:00 LST on July 12.

624


Page 32 of 36

Page 33 of 36


625 33 ACS Paragon Plus Environment


626 627 628 629 630 631

Figure 4. Diurnal cycles of spatially averaged PM2.5 concentrations by species. The left column shows the diurnal cycle of spatially averaged PM2.5 (μg m-3) for CONTROL, COOL_WALL, and COOL_ROOF. The right column shows the differences in PM2.5 species for COOL_WALL – CONTROL and COOL_ROOF – CONTROL and the differences if ventilation effect is excluded. Values represent spatial averages in Los Angeles County (i.e., shown in Figure S3b) for urban grid cells from 00:00 LST on July 3 to 00:00 LST on July 12. Note that vertical axis ranges vary for each species.

632


Page 34 of 36

Page 35 of 36


633 634 635 636 637 638 639

Figure 5. Scatter plots showing fractional increase in EC concentrations induced by cool walls and cool roofs versus (a) fractional reduction in PBL height and (b) fractional reduction in 10-meter wind speed. The value on each dot represents the hour of day (e.g., 9 = 09:00 LST). Least-squares linear regressions and corresponding coefficients of determination (R2) are also shown. Values represent spatial averages in Los Angeles County (i.e., shown in Figure S3b) for urban grid cells from 00:00 LST on July 3 to 00:00 LST on July 12.



ACS Paragon Plus Environment

Page 36 of 36

Investigating the urban air quality effects of cool walls and cool roofs in

Recommend Documents